Global constructive work is associated with ventricular arrhythmias after cardiac resynchronization therapy

Abstract

Aims

Non-invasive left ventricular (LV) pressure–strain loops provide a novel method for quantifying myocardial work by incorporating LV pressure in measurements of myocardial deformation. Early studies suggest that myocardial work parameters such as global constructive work (GCW) could be useful and reliable in arrhythmia prediction, particularly in patients undergoing cardiac resynchronization therapy (CRT). The aim of this study was to evaluate whether the magnitude of GCW was associated with the occurrence of ventricular arrhythmias in patients after CRT.

Methods and results

Patients on guideline-recommended treatment with a CRT defibrillator (CRT-D) were evaluated by 2D speckle-tracking echocardiography including measurements of GCW at least 6 months after implantation. The primary outcome was a composite of appropriate defibrillator therapy and sustained ventricular arrhythmia under the monitor zone. A total of 162 patients [mean age 66 years (±10), 122 males (75%)] were included. Sixteen (10%) patients experienced the primary outcome during a median follow-up of 18 months (interquartile range: 12–25) after the performance of index echocardiography. Patients with a below-median GCW (<1473 mmHg%) had a hazard ratio (HR) for the outcome of 8.14 [95% confidence interval (CI): 1.83–36.08], P = 0.006 compared with patients above the median in a univariate model and remained an independent predictor after multivariate adjustment for the estimated glomerular filtration rate and QRS duration [HR 4.75 (95% CI: 1.01–22.28), P < 0.05].

Conclusion

In patients treated with CRT-D, a GCW below median level was associated with a five-fold increase in the risk of ventricular arrhythmias.

Graphical Abstract

Open in new tabDownload slide

See the editorial comment for this article ‘Assessing left ventricular myocardial work and the risk for malignant arrhythmias: does it work?’, by E. Donal et al., https://doi.org/10.1093/ehjci/jead198.

Introduction

Cardiac resynchronization therapy (CRT) improves cardiac function and survival in properly selected heart failure patients.^1–3 In addition, CRT has an anti-arrhythmic effect that may be related to left ventricular (LV) remodelling and reduction in mechanical dyssynchrony.^4–6 However, more effective methods to predict the occurrence of ventricular arrhythmias (VA) after CRT are lacking. Despite a low sensitivity for detecting arrhythmias, left ventricular ejection fraction (LVEF) is still the most used parameter.⁷

Among several methods proposed, LV mechanical dispersion assessed by 2D speckle tracking has shown some promise in the prediction of arrhythmias.⁸ The presence of high mechanical dispersion as assessed by both radial and longitudinal time-to-peak strain analyses after CRT implantation has been associated with an increased risk of VA.⁹ Studies have demonstrated that persistently high mechanical dispersion after CRT may increase the risk of VA more than two-fold, with an even higher risk of an unfavourable outcome if there is a worsening of mechanical dispersion.⁹

Methods to evaluate mechanical dispersion by 2D speckle-tracking strain analysis are limited by their sensitivity to changes in loading conditions, which affects diagnostic accuracy.^10,11 Global constructive work (GCW), a new potentially useful echocardiographic marker for risk stratification after CRT implantation, accounts for changes in afterload through the incorporation of strain in relation to non-invasively measured blood pressure.^12,13 A high GCW has been associated with a favourable outcome, but the potential predictive value for the development of VA is currently unknown.^14,15

The aim of this study was to evaluate whether GCW is independently associated with VA in patients after CRT defibrillator (CRT-D) implantation and also evaluate whether it significantly adds to risk models including other already established predictors.

Methods

Study population

Patients who had undergone guideline-recommended treatment with a CRT-D were included. All patients were followed up in the out-patient pacemaker and implantable cardioverter defibrillator (ICD) Clinic at Copenhagen University Hospital, Rigshospitalet, Copenhagen, Denmark. The inclusion criteria were: CRT-D implanted in accordance with current guidelines, an available echocardiogram performed at least 6 months after CRT-D implantation, i.e. when most reverse remodelling is expected to have occurred.^16,17 Images were required to be of sufficient quality to perform 2D speckle-tracking analysis and with available brachial blood pressure measurement at the time of the echocardiography. Patients were excluded if they had significant primary valve disease, were revascularized within 3 months of the echocardiography, or developed an acute coronary syndrome during follow-up.

Outcome

The primary outcome was a composite outcome of any appropriate ICD therapy [shock therapy or anti-tachycardia pacing (ATP)], sustained ventricular tachycardia (VT) >30 s. In secondary analysis, all-cause death was added to the primary outcome. Patients were followed for up to 3 years after the baseline echocardiography or until the occurrence of one of the outcomes of interest. All data on arrhythmias were collected through ICD interrogations during hospitalization, out-patient visits, or during transmission from home monitoring (HM) systems. The HM system is programmed to report biannually or within 24 h of shock therapy and HM is offered to all patients treated with CRT-D in Denmark. Death events were collected using the electronic patient files. All outcomes were adjudicated by a steering committee of experienced electrophysiologists blinded to the echocardiographic data.

Echocardiography

All echocardiographies were performed using a GE Healthcare Vivid E9 system. Frame rates were defined per protocol to be 50–90 frames/s on greyscale images and >130 frames/s with tissue Doppler imaging (TDI). All standard images were recorded with three cine loops. Measurements of end-systolic atrial volume, tricuspid annular plane systolic excursion (TAPSE), and LVEF using Simpson’s Biplane Method were performed. The timing for aortic- and mitral valve opening and closing was determined by TDI m-mode through the anterior leaflet of the mitral valve in the four-chamber view.

Myocardial work

The blood pressure was measured using a brachial artery cuff in a sitting position for all patients immediately preceding the echocardiography. In patients without valve disease, peak arterial pressure shows good correlation with peak systolic LV pressure.¹³ Apical two-, three-, and four-chamber views focused on the LV were used for the analyses of longitudinal strain. Automated software estimated a systolic LV pressure curve using valvular event timings and systolic LV pressure.^14,18,19 End systole was defined as the time of aortic valve closure. Strain rate functions were multiplied by the LV pressure–strain function to obtain a function of LV power—Figure 1A. The integral of the LV power function produces myocardial work as a function of time. GCW is defined as the myocardial work within the area from mitral valve closure to mitral valve opening.²⁰ Thus, GCW is the myocardial work contributing to the LV ejection period during systole (segmental shortening of the myocytes during systole and the lengthening of the myocytes during isovolumic relaxation)—Figure 1A. LV dyssynchrony, an early or late activation of segments causing differential contraction times between myocardial segments, is highly associated with a marked imbalance in segmental work distribution. Thus, severe dyssynchrony causes a significant decrease in GCW¹² visualized in Figure 1E.

Ethics and approvals

The study was approved by the Danish Data Protection Agency as well as the Regional Danish Committee on Health Research Ethics and was conducted in accordance with the principles outlined in the Declaration of Helsinki.

Statistical analyses

Continuous data are presented as mean and standard deviation or median and interquartile range (IQR), as appropriate. Categorical data are expressed as numbers and percentages. Kaplan–Meier survival curves were calculated for freedom from outcomes in patient groups with above- and below-median GCW and mechanical dispersion. Patients with above- and below-median GCW and mechanical dispersion will henceforth be referred to as patients with high/low GCW and mechanical dispersion, respectively, throughout the article. Subdistribution of death as a competing risk was assessed by the method proposed by Fine and Gray.²¹ Cox regression analysis was used in the patient population to identify the predictors of the outcomes. Variables used in the multivariate model were selected by using stepwise backwards regression. Receiver-operator characteristics (ROC) curves were used to determine optimal cut-offs from the point closest to the upper left corner to enable GCW and mechanical dispersion to identify the outcomes. The predictive strength of GCW and time-to-peak dyssynchrony was compared using −2 log-likelihood statistics. A predefined cut-off for time-peak dyssynchrony of 70 ms was used based on previous findings in other cohorts.⁹ Additional analyses were performed below or above the median septal-lateral work ratio and the primary outcome, as this parameter is known to have strong predictive value for the benefit of CRT treatment.²² All statistical analyses were performed using a standard statistical software program (R statistical software, version 4.1.0).

Results

Clinical characteristics of the cohort

A total of 166 eligible patients were identified during the period from February to October 2021. After the exclusion of 4 patients due to poor image quality of the echocardiographies, 162 patients were included in the study. Baseline characteristics at the time of inclusion are presented in Table 1. The mean age of the patients was 66 ± 10 years, 122 (75%) were male, and 68 (42%) had ischaemic heart disease. The mean LVEF was 40 ± 11%. The patients were generally well treated, with 90% receiving a beta-blocker, 89% receiving an angiotensin-converting enzyme-, angiotensin II-, or an angiotensin receptor-neprilysin inhibitor, and 62% receiving mineralocorticoid receptor antagonists.

A total of 18% of the patients were treated with CRT-D on a pacing indication, and 82% were treated because of the presence of a primary bundle branch block. In total, 14% had atrial fibrillation, and the mean biventricular pacing percentage was 96. The median time between CRT-D implantation and index echocardiography was 18 months (12–25).

The median GCW was 1473 mmHg% (IQR: 997–1774). Stratification according to high and low GCW is presented in Table 1. Patients with high GCW had a shorter QRS duration of 142 ± 24 vs. 159 ± 24 ms (P < 0.001), were less likely to be male, 65 vs. 85% (P = 0.006), and had better cardiac function by conventional echocardiographic measures (LVEF, TAPSE, and end-systolic and end-diastolic LV volumes) compared with those with low GCW.

Follow-up and outcomes

The median follow-up time was 18 months (IQR: 12–25). Of the 162 patients, 16 (10%) experienced the primary outcome, 7 patients (4%) had appropriate shock therapy, and 9 patients (6%) had appropriate ATP therapy. None of the patients experienced sustained VT or ventricular fibrillation (VF) that were not appropriately terminated by the ICD. In total, 23 (14%) patients experienced the secondary outcome, and 8 (5%) of them died.

GCW levels and outcomes

Patients with low GCW after CRT-D implantation were at a higher risk of ventricular arrhythmias when compared with those with high GCW; hazard ratio (HR) 8.1 [95% confidence interval (CI) 1.8–36.1], P = 0.006 (Figure 2). Patients with low GCW were also at a high risk of VA and death with HR 5.7 [1.9–16.8], P = 0.002. The risk of VA was not modified by precluding competing events of death between the two groups (Gray’s test P = 0.25). A low GCW level was incrementally associated with a poor outcome. For every 100 mmHg% decrease in GCW, the risk of VA increased by 13% (1–28), P = 0.04.

ROC curve analysis showed that the optimal cut-off value for the prediction of VA was 1333 mmHg% for GCW (Figure 3) with area under the curve (AUC) of 0.76 (0.64–0.87). The optimal cut-off for prediction of VA and death was 1248 mmHg% with an AUC of 0.75 (0.65–0.86). Comparing multivariate risk models with known predictors of VA including LVEF <35%, ischaemic heart disease, QRS duration, New York Heart Association function class, and mechanical dispersion; GCW significantly added to the association with VA (P = 0.02 for differences).

Among covariates, estimated glomerular filtration rate (eGFR), QRS duration, and LVEF were significantly associated with outcome in univariate analysis (Figure 4). Stepwise backwards regression determined that the model that best explained the data was adjusted for eGFR and QRS duration. GCW remained independently associated with outcome with an almost five-fold increase in the risk of VA in patients with low compared with patients with a high GCW after multivariate adjustment [HR 4.8 (1.01–22.3), P < 0.05]. After addition of death to the arrhythmia outcome the independent association remained significant [HR 3.35 (1.08–10.40), P = 0.04]—Figure 5.

Comparison to mechanical dispersion

Time-to-peak dyssynchrony below and above 70 ms was not significantly associated with outcome after adjustment in the multivariate analysis 1.52 (0.44–5.24), P = 0.51 (Figure 4).

Septal-lateral work ratio

Septal-lateral constructive work ratio was not associated with the primary outcome: HR 1.33 (0.49–3.58), P = 0.57 in an unadjusted model and HR 1.06 (0.39–2.89), P = 0.90 in a multivariate adjusted model.

Discussion

The present study is the first to demonstrate the association between GCW levels after CRT-D implantation and the risk of VA. Low GCW >6 months after implantation was independently associated with an increased risk of VA as well as the combination of VA and death within 18 months of follow-up. Furthermore, GCW was superior to LVEF and time-to-peak dyssynchrony for identifying patients with a poor prognosis.

Prediction of arrhythmia after CRT

CRT is aimed at reducing LV electromechanical dyssynchrony, hereby reversing LV remodelling and improving cardiac contraction and efficiency.²³ Landmark clinical CRT trials have demonstrated that CRT reduces the risk of VA and sudden cardiac death by 36%;¹⁶ however, the mechanisms behind the anti-arrhythmic effect of CRT are not fully established, and there is no consensus whether it can be ascribed to mechanical or electrical effects. The higher LVEF after CRT implantation, the lower the risk of VA.^7,24 Studies have demonstrated a 76% lower risk of VA in patients with near-normalized LVEF after CRT implantation compared with those without normalization during 2.2 years of follow-up [HR 0.24 (0.07–0.82), P = 0.02].⁷

Reduction in LV volumes that follow successful resynchronization is accompanied by a reduction in wall tension and changes in neurohormonal activation which is thought to decrease arrhythmogenicity of the myocardium.²⁵ In addition, a more synchronized electrical activation of the LV likely reduces the variation in refractoriness, and this is regarded as reducing the potential for re-entry arrhythmias²⁶ (see Supplementary data online, Figure S1).

Tools to evaluate the risk of arrhythmias after CRT implantation are sparse. Assessment of mechanical dyssynchrony using 2D speckle-tracking strain analysis has been suggested to be useful, especially mechanical dispersion based on time-to-peak analysis.²⁷ In a study by Haugaa et al.,⁹ persistent dyssynchrony after CRT was associated with a 2.5-fold increased risk of VA relative to the risk in patients without dyssynchrony. This risk is even further increased with worsened dyssynchrony after CRT. Similarly, Kutyifa et al.⁶ have demonstrated that a 15% decrease in LV dyssynchrony after CRT implantation is associated with a 70% reduction in absolute risk of VT/VF.

Myocardial work

The current study showed improved risk prediction of VA when using GCW compared with time-to-peak analysis. A major advantage of myocardial work parameters compared with other markers of myocardial function is the incorporation of systemic blood pressure as a proxy for afterload. LVEF and strain are sensitive to changes in afterload and may lead to false conclusions of decreased contractility if the systolic blood pressure is elevated. Afterload is, however, taken into account when calculating GCW.¹³ The work performed in the different segments of the heart is correlated with differences in the blood flow and oxygen demand which contributes to remodelling of the LV. This makes GCW a potential method to examine the haemodynamic impact of dyssynchrony to monitor the incidence of arrhythmias.²⁸ Previous studies have demonstrated that the estimation of GCW prior to CRT implantation is an independent predictor of CRT outcome.^14,15 In a study by Galli et al.,¹⁵ the presence of lower GCW was associated with a less favourable response to CRT, GCW <1057 mmHg% identified 87% of non-responders with a positive predictive value of 88%. Another study by Galli et al.¹⁴ of 166 CRT patients showed that those with persistent GCW <888 mmHg% had a nearly five-fold increased risk of cardiac death.

In the current study, GCW was evaluated >6 months after CRT-D implantation to ascertain LV reverse remodelling. The GCW values in the current study reflect that patients had already benefitted from CRT at the time of the index echocardiogram. CRT-D patients with persistently low GCW had a five-fold increased risk of VA and a three-fold increased risk of VA and death during a median of 18 months of follow-up.

Clinical perspective

The current study suggests that the assessment of GCW after reverse remodelling in response to CRT-D implantation may be useful in risk stratification for VA. Few parameters are available to establish the prognosis in CRT-D treated patients with heart failure and GCW may be useful for this purpose. A dichotomous predictor (GCW ≤ 1473) is difficult to establish, but the risk of arrhythmia increases linearly with decreasing values (13% pr 100 mmHg), and GCW may be useful to improve the prognostic expectations in the future.

Limitations

Patients were included retrospectively in a non-randomized setting. Index echocardiography was not performed immediately after CRT-D implantation for all patients. However, all patients had the index echocardiography performed at least 6 months after CRT-D implantation to ensure LV reverse remodelling at least partly. An inherent limitation of this method is the use of brachial blood pressure, as a non-invasive measurement, for the estimation of LV pressure. Therefore, GCW is less accurate when evaluating patients with conditions that cause large differences in systemic- and LV pressure (e.g. aortic stenosis). Those patients were therefore excluded in the current study. Pre-implantation data were not investigated in this study. A future prospective study is needed that investigates how pre-implantation GCW and changes in GCW after remodelling predict arrhythmic events. This study also did not include data on wall viability, which may be valuable for risk prediction.

Conclusion

In patients with heart failure treated with CRT-D, GCW below the median level over 6 months after implantation was associated with a five-fold increase in the risk of VA. GCW was demonstrated to provide improved risk prediction for arrhythmias after CRT-D relative to mechanical dispersion.

Supplementary data

Supplementary data are available at European Heart Journal - Cardiovascular Imaging online.

Funding

This work was supported by the Danish Heart Foundation, Snedkermester Sophus Jacobsen og hustru Astrid Jacobsens Fond, the Hartmann Foundation, and the Novo Nordisk Foundation. The funders were not involved in planning the study or in the decision to publish.

Data availability

Data can be made available from the corresponding author upon reasonable request and if such request is in accordance with the Danish Data Protection Act and the General Data Protection Regulation.

References

Author notes